In multiagent team settings, the agents are often given a protocol for coordinating their actions. When such a protocol is not available, agents must engage in ad hoc teamwork to effectively cooperate with one another. A fully general ad hoc team agent needs to be capable of collaborating with a wide range of potential teammates on a varying set of joint tasks. This paper extends previous research in a new direction with the introduction of an efficient method for reasoning about the value of information.  Then, we show how previous theoretical results can aid ad hoc agents in a set of testbed pursuit domains.

Agents in highly dynamic adversarial domains, such as RTS games, must continually make time-critical decisions to adapt their behaviour to the changing environment. In such a context, the planning agent must consider his opponent's actions as uncontrollable, or at best influenceable. In general nondeterministic domains where there is no clear turn-taking protocol, most heuristic search methods to date do not explicitly reason about the opponent's actions when guiding the state space exploration towards goal or high-reward states. In contrast, we are investigating a domain-independent heuristic planning approach which reasons about the dynamics and uncontrollability of the opponent's behaviours in order to provide better guidance to the search process of the planner. Our planner takes as input the opponent's behaviours recognized by a plan recognition module and uses them to identify opponent's actions that lead to low-utility projected states. We believe such explicit heuristic reasoning about the potential behaviours of the opponent is crucial when planning in adversarial domains, yet is missing in today's planning approaches.

Transfer learning as a new machine learning paradigm has gained increasing attention lately. In situations where the training data in a target domain are not sufficient to learn predictive models effectively, transfer learning leverages auxiliary source data from other related source domains for learning. While most of the existing works in this area only focused on using the source data with the same structure as the target data, in this paper, we push this boundary further by proposing a heterogeneous transfer learning framework for knowledge transfer between text and images. We observe that for a target-domain classification problem, some annotated images can be found on many social Web sites, which can serve as a bridge to transfer knowledge from the abundant text documents available over the Web. A key question is how to effectively transfer the knowledge in the source data even though the text can be arbitrarily found. Our solution is to enrich the representation of the target images with semantic concepts extracted from the auxiliary source data through a novel matrix factorization method. By using the latent semantic features generated by the auxiliary data, we are able to build a better integrated image classifier. We empirically demonstrate the effectiveness of our algorithm on the Caltech-256 image dataset.

Effective access to knowledge within large declarative memory stores is one challenge in the development and understanding of long-living, generally intelligent agents. We focus on a sub-component of this problem: given a large store of knowledge, how should an agent's task-independent memory mechanism respond to an ambiguous cue, one that pertains to multiple previously encoded memories. A large body of cognitive modeling work suggests that human memory retrievals are biased in part by the recency and frequency of past memory access. In this paper, we evaluate the functional benefit of a set of memory retrieval heuristics that incorporate these biases, in the context of the word sense disambiguation task, in which an agent must identify the most appropriate word meaning in response to an ambiguous linguistic cue. In addition, we develop methods to integrate these retrieval biases within a task-independent declarative memory system implemented in the Soar cognitive architecture and evaluate their effectiveness and efficiency in three commonly used semantic concordances.

We consider a new class of planning problems involving a set of non-negative real variables, and a set of non-deterministic actions that increase or decrease the values of these variables by some arbitrary amount. The formulas specifying the initial state, goal state, or action preconditions can only assert whether certain variables are equal to zero or not. Assuming that the state of the variables is fully observable, we obtain two results. First, the solution to the problem can be expressed as a policy mapping qualitative states into actions, where a qualitative state includes a Boolean variable for each original variable, indicating whether its value is zero or not. Second, testing whether any such policy, that may express nested loops of actions, is a solution to the problem, can be determined in time that is polynomial in the qualitative state space, which is much smaller than the original infinite state space. We also report experimental results using a simple generate-and-test planner to illustrate these findings.

The conservation of wildlife corridors between existing habitat preserves is important for combating the effects of habitat loss and fragmentation facing species of concern. We introduce the Steiner Multigraph Problem to model the problem of minimum-cost wildlife corridor design for multiple species with different landscape requirements. This problem can also model other analogous settings in wireless and social networks. As a generalization of Steiner forest, the goal is to find a minimum-cost subgraph that connects multiple sets of terminals. In contrast to Steiner forest, each set of terminals can only be connected via a subset of the nodes. Generalizing Steiner forest in this way makes the problem NP-hard even when restricted to two pairs of terminals. However, we show that if the node subsets have a nested structure, the problem admits a fixed-parameter tractable algorithm in the number of terminals. We successfully test exact and heuristic solution approaches on a wildlife corridor instance for wolverines and lynx in western Montana, showing that though the problem is computationally hard, heuristics perform well, and provably optimal solutions can still be obtained.

MAPP has been previously shown as a state-of-the-art multi-agent path planning algorithm on criteria including scalability and success ratio (i.e., percentage of solved units) on realistic game maps. MAPP further provides a formal characterization of problems it can solve, and low-polynomial upper bounds on the resources required. However, until now, MAPP's solution quality had not been extensively analyzed. In this work we empirically analyze the quality of MAPP's solutions, using multiple quality criteria such as the total travel distance, the makespan and the sum of actions (including move and wait actions). We also introduce enhancements that improve MAPP's solution quality significantly. For example, the sum of actions is cut to half on average. The improved MAPP is competitive in terms of solution quality with FAR and WHCA*, two successful algorithms from the literature, and maintains its advantages on different performance criteria, such as scalability, success ratio, and ability to tell apriori if it will succeed in the instance at hand. As optimal algorithms have limited scalability, evaluating the quality of the solutions provided by suboptimal algorithms is another important topic. Using lower bounds of optimal values, we show that MAPP's solutions have a reasonable quality. For example, MAPP's total travel distance is on average 19% longer than a lower bound on the optimal value.

Sustainable resource management in many domains presents large continuous stochastic optimization problems, which can often be modeled as Markov decision processes (MDPs). To solve such large MDPs, we identify and leverage linearity in state and action sets that is common in resource management. In particular, we introduce linear dynamic programs (LDPs) that generalize resource management problems and partially observable MDPs (POMDPs). We show that the LDP framework makes it possible to adapt point-based methods--the state of the art in solving POMDPs--to solving LDPs. The experimental results demonstrate the efficiency of this approach in managing the water level of a river reservoir. Finally, we discuss the relationship with dual dynamic programming, a method used to optimize hydroelectric systems.

Most of the previous work on brain-computer interfaces (BCIs) exploiting the P300 in electroencephalography (EEG) has focused on low-level signal processing algorithms such as feature extraction and classification methods. Although a significant improvement has been made in the past, the accuracy of detecting P300 is limited by the inherently low signal-to-noise ratio in EEGs. In this paper, we present a systematic approach to optimize the interface using partially observable Markov decision processes (POMDPs). Through experiments involving human subjects, we show the P300 speller system that is optimized using the POMDP achieves a significant performance improvement in terms of the communication bandwidth in the interaction.

Plants can have positive effects on each other in numerous ways, including protection from harsh environmental conditions. This phenomenon, known as facilitation, occurs in water-stressed environments when shade from larger shrubs protects smaller annuals from harsh sun, enabling them to exist on scarce water. The topic of this paper is a model of this phenomenon that allows search algorithms to find residential landscape designs that incorporate facilitation to conserve water. This model is based in botany; it captures the growth requirements of real plant species in a fitness function, but also includes a penalty term in that function that encourages facilitative interactions with other plants on the landscape. To evaluate the effectiveness of this approach, two search strategies--simulated annealing and agent-based search--were applied to models of different collections of simulated plant types and landscapes with different light distributions. These two search strategies produced landscape designs with different spatial distributions of the larger plants. All designs exhibited facilitation and lower water use than designs where facilitation was not included.